Google’s parent company, Alphabet, has just experienced its worst day on the stock market since Google went public in 2004. Its shares fell around 12.6% on Thursday, 29 October, following its latest quarterly results. Despite reporting strong revenue growth, the company recorded negative free cash flow and a sharp increase in spending on property and equipment as it continues expanding its AI infrastructure.
Other Big Tech companies are making similarly large investments in AI development, data centres and computing power. Some see this spending as an opportunity to transform businesses and create new revenue streams, while others are concerned about the costs and whether the financial returns will justify them. For the companies writing the cheques, having the most powerful models or the biggest infrastructure budget may not be enough. Investors want to see what all that spending can deliver and businesses need to work out where AI can make a meaningful difference.
Investors Would Like To See The Money Come Back
The recent decline in Google’s market value reflects changing expectations around Big Tech’s AI ambitions, according to Olga Kokhan, Founder and CEO of Tinkogroup. She believes companies are entering a more disciplined phase of AI adoption, where sustainable business value carries more weight than simply keeping up with competitors.
“Simply adopting AI to keep up with competitors is not a strategy,” Kokhan says.
Rahul Bokde, Founder of Toolspike, sees the spending from a smaller software company’s perspective. Falling model costs are making increasingly capable AI tools more accessible, allowing businesses to build products without carrying the full cost of developing the underlying technology themselves. However, he believes investors will want more than the promise of increasingly capable models.
“If a capability gets cheaper every few months, the return on the data centres cannot come from the model alone,” Bokde says.
For companies such as Google, this could mean demonstrating how AI is generating revenue through products people already use, including search, cloud services and workplace software.
Wiktor Stachowski, Founder of LeadScout, has seen the benefits of falling AI costs first-hand. As a solo founder using AI APIs, he says cheaper model calls have made it possible to build low-cost research tools for sales teams. Larger companies’ infrastructure investment is helping smaller businesses develop products without having to build their own computing infrastructure.
“In practice, those billions in infrastructure are already subsidising thousands of small products,” Stachowski says.
He expects investors to become more focused on the relationship between AI spending and the revenue it generates. As he puts it, “I expect 2027 to be the year investors stop rewarding ‘AI spend’ as a headline and start asking for revenue per dollar of compute.”
More from Artificial Intelligence
- What Happens When Your AI Assistant Knows More About You Than Your Closest Friends?
- The Biggest AI Mistakes Small Businesses Are Making
- The Future Of Incident Management Depends On Humans And AI Working Together
- Is English Actually Easier For AI To Understand Because Of Its Syntax?
- The AI Asset Hiding At British Companies: Years Of Ordinary Work
- What Is AI Inference?
- Is AI Making It Easier To Build A Startup, Or Just More Difficult To Stand Out? The Experts Weigh In
- Beyond Headcount: How AI Is Changing The Way Businesses Plan Their Workforces
Topping The Leaderboard Does Not Get The Work Done
While the cost of AI infrastructure and model development attracts plenty of attention, Jonathan Lujan, CEO of MatrixCare, believes its economic impact should be measured through the productivity it creates for organisations and the people using it. He points to the potential for businesses to serve more customers without increasing costs at the same rate, while freeing skilled employees to focus on higher-value work.
Healthcare could provide an important example as Lujan hopes AI will help caregivers and other healthcare professionals improve business operations and patient care, rather than just adding another layer of technology. He also believes some of AI’s greatest value could go to the businesses using it, rather than the companies developing the underlying models.
Royce Cheng, Chief Product and Technology Officer at Third Way Health, sees a gap between what AI can achieve in testing environments and what businesses need from it in practice. He explains that AI was previously judged heavily on benchmark performance, but buyers are now looking for results that work with their own data, processes and industry knowledge.
Cheng argues that businesses can’t simply onboard an AI system like a new employee, as many existing workflows weren’t designed for always-on agents. Successful adoption requires more than choosing a capable model, with governance, staff training and change management all playing a part.
Railways, Bubbles And Whoever Is Holding The Debt
Not everyone is convinced that Big Tech should be pulling back on AI spending. James Buckley-Thorp, Founder and CEO of Atlian; has a more optimistic view, comparing today’s investment to the development of the railways. Railway infrastructure required enormous upfront investment, with its wider economic benefits becoming apparent over time. Buckley-Thorp believes AI could follow a similar path, with some of its biggest gains emerging in traditional industries such as insurance and construction.
These sectors still deal with substantial amounts of paperwork and administrative work, creating opportunities for AI to reduce manual processes and change how certain jobs are carried out. His view is that companies should continue investing rather than retreating in response to market uncertainty.
“Spend more, not less,” Buckley-Thorp says.
However, the financial risks extend beyond the technology companies making the investments. Nilesh Chavda, Digital Strategy Lead at VIP Apps Consulting, points to the exposure of banks and private credit providers financing the expansion of AI infrastructure. Data centres are being funded through debt, project finance and other financial arrangements, with some of that investment relying on expectations of future demand for computing capacity.
Chavda warns that lenders could face problems if the demand or revenue needed to support these projects fails to materialise. There’s also the risk that hardware becomes outdated faster than the software businesses built around it can generate returns. For financial services, he sees opportunities in AI through more specialised applications, including automating industry-specific processes and connecting existing systems through low-code integration.
Big Tech’s AI Spending Now Has To Show Its Workings
Big Tech’s AI investments are creating opportunities for smaller businesses, but they’re exposing technology companies and their financial backers to substantial costs. The next stage will involve proving that the infrastructure can generate lasting commercial returns, rather than simply adding to the industry’s spending figures.
Our Experts:
- Jonathan Lujan: CEO of MatrixCare
- Royce Cheng: Chief Product and Technology Officer at Third Way Health
- Rahul Bokde: Founder of Toolspike
- Olga Kokhan: Founder and CEO of Tinkogroup
- James Buckley-Thorp: Founder and CEO of Atlian
- Wiktor Stachowski: Founder of LeadScout
- Nilesh Chavda: Digital Strategy Lead at VIP Apps Consulting
Jonathan Lujan, CEO of MatrixCare
“Billions are being invested in AI infrastructure, and understandably, investors want to know when those investments will pay off.”
“The more important measure of AI’s economic impact will be how much productive capacity it creates across the broader economy and especially with the end user.”
“Does it allow organisations to serve more customers without proportionally increasing costs? Does it free up skilled employees to focus on higher-value work?”
“There will be significant returns for the companies that invest in AI, but the companies and people that capture the greatest value from AI may not be those selling the underlying technology.”
“In healthcare, I hope that it’s the end users, the caregivers who actually improve their business models to provide better care in a better way for patients that get the biggest and best return.”
Royce Cheng, Chief Product & Technology Officer at Third Way Health
“Up until the last few months, frontier AI has largely been measured against benchmarks, and you saw different tech companies jostling for the top spot of benchmark leaderboards.”
“Now, AI buyers are asking for practical results, which require something that benchmarks lack: the data, processes, and knowledge specific to the buyer’s organisation.”
“And it isn’t as simple as ‘onboarding AI as you would an employee’, because existing ways of working were never designed for always-on, exhaust-all-options AI agents.”
“Safely harnessing AI is fundamentally different from extending the human workforce, and requires governance, training, and change management far beyond what is encapsulated in an AI model.”
Rahul Bokde, Founder of Toolspike
“From where a small software company sits, Big Tech’s AI spending is already paying off for customers: capable models now cost a cent or two per task, and prices have been falling.”
“That is exactly what worries investors. If a capability gets cheaper every few months, the return on the data centres cannot come from the model alone.”
“It has to come from volume and from owning the customer relationship, through search, cloud contracts and office software.”
“I expect investors to stop rewarding capital spending on its own and to ask for revenue tied to specific AI products.”
Olga Kokhan, Founder & CEO of Tinkogroup
“The recent decline in Google’s market value reflects a growing shift in how investors evaluate Big Tech’s AI ambitions. The key of this issue is which companies can turn investment in AI into sustainable business value.”
“I believe we’re entering a more disciplined phase of AI adoption. Building infrastructure and developing powerful models require enormous capital, but the long-term winners will be companies that can translate these capabilities into measurable productivity gains, lower operating costs, and new revenue streams.”
“This will also put pressure on businesses to rethink how they approach AI implementation. Simply adopting AI to keep up with competitors is not a strategy.”
“Companies need to identify where automation genuinely improves operations and where human expertise remains essential.”
James Buckley-Thorp, Founder and CEO of Atlian
“Google’s share price wobble is the best thing that could have happened to AI, because it’s shaking out the tourists.”
“If you’re panicking about Big Tech spending billions on compute, you’d have panicked about the railways too. Of course the spreadsheets look insane halfway through. That’s what building the future looks like.”
“The real money won’t show up in Silicon Valley earnings calls first. It’ll show up in dusty, old-fashioned industries like insurance and construction, where AI is already wiping out weeks of paperwork and, frankly, a fair few pointless jobs with it.”
“Good. The only genuine risk here is Britain bottling it and regulating AI into the ground before we’ve even started. Spend more, not less.”
Wiktor Stachowski, Founder of LeadScout
“As a solo founder whose product runs entirely on AI APIs, I see the other end of Big Tech’s spending. The cost per model call keeps falling, and that’s the only reason a small company like mine can sell AI research to sales teams for a few euros a month.”
“In practice, those billions in infrastructure are already subsidising thousands of small products.”
“The question investors are starting to ask is who captures that value. If model access becomes a cheap commodity, the margin moves to whoever owns the customer and the specific workflow, not to whoever trains the biggest model.”
“I expect 2027 to be the year investors stop rewarding ‘AI spend’ as a headline and start asking for revenue per dollar of compute. Google’s drop looks like an early sign of that shift.”
Nilesh Chavda, Digital Strategy Lead at VIP Apps Consulting
“The tech market sell-off exposes a growing risk for financial services: banks and private credit funds are heavily funding hyperscaler CapEx through data centre debt, project finance, and off-balance-sheet vehicles.”
“By underwriting physical infrastructure on speculative future demand, financial institutions are absorbing significant credit risk in a potential AI bubble.”
“High-performance hardware quickly becomes obsolete, leaving lenders exposed if asset values fall faster than the software revenues brought in to service the debt.”
“Financial institutions do not need bigger, all-purpose AI models.”
“Sustainable returns in financial services will come from domain-specific process orchestration and low-code integration, turning heavy capital expenditure into lean, compliant, and efficient daily operations across underwriting, risk, and portfolio management.”
